Toward imaging biomarkers for glycosaminoglycans

Abstract

Advances in the diagnosis and treatment of cartilage degeneration will be accelerated with the availability of validated biomarkers that reveal the features relevant to the health of cartilage. Using the delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) technique for evaluating tissue glycosaminoglycan as a case study, I review the types of evidence needed to validate imaging (or other) biomarkers. In addition, I present discussions about face validity and technical validity and offer a review of emerging data that provide pathophysiologic validity. Examples of such data include evidence that glycosaminoglycan content is restored after an injury-induced loss and evidence suggesting that dGEMRIC can indicate when it is too late for protective (load-modifying) surgery. These and other data suggest that new imaging biomarkers may indeed be able to provide a state-of-cartilage proxy that can be of use in the diagnosis and staging of disease.

Fig. 1

The biomarker development process. The broad stages of biomarker development include establishment of face validity, technical validity, and pathophysiologic validity. Development generally involves iterating among these, as additional pathophysiologic insights are gained.

Fig. 2

dGEMRIC (A) and T1rho (B) versus macromolecular concentration for glycosaminoglycan (GAG) and collagen suspensions. dGEMRIC shows an approximately linear dependence on GAG, with no dependence on collagen; T1rho shows approximately exponential dependencies on both GAG and collagen. The thicker regions of the trend lines indicate the regions corresponding with the range of values found in native cartilage: GAG ranges from approximately 7% at the high end of normal to 0% in disease, and collagen ranges from approximately 15% to 25%. Gd(DTPA)^2- = gadolinium diethylenetriamine pentaacetic acid. (Reproduced, with modification, from: Menezes NM, Gray ML, Hartke JR, Burstein D. T2 and T1rho MRI in articular cartilage systems. Magn Reson Med. 2004;51:503-9. Reprinted with permission.)

Fig. 3

The spatial distribution of T1 in the presence of gadolinium diethylenetriamine pentaacetic acid (Gd[DTPA]^2-) in a preoperative clinical image is similar to that seen for the same tissue excised and imaged postoperatively, and both provide a similar qualitative impression of glycosaminoglycan (GAG) distribution as the corresponding toluidine-blue stained histological section. The arrows indicate regions for comparison. The plot on the right-hand side compares dGEMRIC T1 values measured clinically (preoperatively) with those measured in vitro (postoperatively) for the same regions of interest in the same joint in four patients. Each patient is shown with a different symbol. To allow comparison, values for each patient are shown normalized to one region of interest in each joint. The high correlation suggests that in vivo T1 images in the presence of Gd(DTPA)^2- provide an accurate assessment of the concentration of GAG relative to other regions in the same tissue. (Reproduced, with modification, from: Bashir A, Gray ML, Hartke J, Burstein D. Nondestructive imaging of human cartilage glycosaminoglycan concentration by MRI. Magn Reson Med. 1999;41:857-65. Reprinted with permission.)

Fig. 4

dGEMRIC images of the right medial compartment from before and after a posterior cruciate ligament (PCL) tear showing a drop from the baseline dGEMRIC index at one and three months after the injury and a return to baseline values by six months. (Reproduced, with modification, from: Young AA, Stanwell P, Williams A, Rohrsheim JA, Parker DA, Giuffre B, Ellis AM. Glycosaminoglycan content of knee cartilage following posterior cruciate ligament rupture demonstrated by delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC). A case report. J Bone Joint Surg Am. 2005;87:2765.)

Fig. 5

The dGEMRIC index, together with subluxation, is a strong predictor of the outcome of osteotomy. A: dGEMRIC index for subjects with hip dysplasia who have undergone periacetabular osteotomy (PAO), grouped by those who had an unsatisfactory result (defined as secondary arthroplasty, increased pain, and/or decreased joint-space width) (n = 10) and those who did not (n = 42). B: Statistical analysis indicated that the preoperative dGEMRIC index and subluxation were significant predictors of early postoperative failure. (Reproduced, with modification, from: Cunningham T, Jessel R, Zurakowski D, Millis MB, Kim YJ. Delayed gadolinium-enhanced magnetic resonance imaging of cartilage to predict early failure of Bernese periacetabular osteotomy for hip dysplasia. J Bone Joint Surg Am. 2006;88:1543-4.)